Person-centered care (PCC) theory holds that actions, emotions, and well-being of individuals with dementia reflect the quality of their interactions with others. Kitwood (1993) emphasized that personhood is the concurrence of internal and external relationships. Enriching social relationships support the individual's “self,” counteracting disabling effects of dementia. Unsupportive relationships threaten the self and diminish individual capacity (Kitwood, 1993). Implementing PCC requires exploring the quality of residents' relationships, informed by knowledge of the individual's history, perceptions, and observation of his/her interactions with others (Stein-Parbury et al., 2012).
Lack of social relationships and perceived lack of support are associated with chronic illness and reduced cognition (Cacioppo, Capitanio, & Cacioppo, 2014). Social engagement is associated with better cognition and quality of life in older adults (Golden, Conroy, & Lawlor, 2009). Greater number and interconnectedness of relationships are associated with more opportunity for engagement (Ashida & Heaney, 2008). Nursing home (NH) residents' relationships with co-residents contribute uniquely to perceived support, yet recent research indicated that residents had few friendships and found forming relationships difficult (Casey, Low, Jeon, & Brodaty, 2015). Socioemotional selectivity theory suggests that older adults seek emotional equilibrium through selecting emotionally rewarding relationships over less rewarding ones, thus retaining supportive relationships and narrowing their social networks (English & Carstensen, 2014). Maintaining a “balanced” ratio of 2.9 positive emotions for every negative emotion was associated with “flourishing” in NH residents, including those with mild cognitive impairment (Meeks, Van Haitsma, Kostiwa, & Murrell, 2012, p. 465).
Social network analysis (SNA) investigates relationships (i.e., ties) between individuals and the effects associated with these relationships (Borgatti, Everett, & Freeman, 2002). Using SNA to investigate co-resident relationships within NHs can inform PCC by: (a) illustrating relational patterns forming the broader social context within which residents experience relationships; and (b) identifying how these influence residents' engagement and perceived support (Casey et al., 2015). The current authors used SNA to describe care staff perceptions of NH residents' social networks and explore associations between network characteristics and resident engagement and social isolation.
Study design and protocol were approved by the University of South Wales Australia Human Research Ethics Committee. The cross-sectional, multiple-method design used purposive sampling in a Sydney NH. Care units were: 42-bed Unit 1, 18-bed dementia-specific unit (DSU), and 34-bed Unit 3. All residents not acutely ill were invited to participate (N = 91). Study consent protocols are explained in detail elsewhere (Casey et al., 2015). Thirty-six residents provided either informed signed consent (n = 4) or verbal assent with legal guardian's informed signed consent (n = 32) to full participation. Fifty-five additional residents were included in network data only (Casey et al., 2015). Informants were six permanent care staff who knew the residents well, including the Recreational Activities Officer, two Assistants in Nursing, and three RNs.
Residents' demographic data (i.e., age, sex, marital status) and medical history (i.e., time in care, dementia diagnosis) were transcribed from medical charts. Ability in activities of daily living (ADL) was assessed using chart data and resident interviews to complete the 17-item Barthel Index (Mahoney & Barthel, 1965). Staff reported resident social engagement using the Multidimensional Observation Scale for Elderly Subjects (MOSES)—Withdrawn Behavior subscale (Helmes, Csapo, & Short, 1987). Residents reported perceived social support using The Friendship Scale (Hawthorne, 2006). Internal consistencies were good and acceptable, respectively (MOSES: Cronbach's alpha = 0.80, n = 36; The Friendship Scale: Cronbach's alpha = 0.76, n = 21). Residents reported objective social support using a version of the 3-item Lubben Social Network Scale-6 (LSNS-6) Friendship subscale (Lubben et al., 2006) adapted for NH use, with multiple timeframe options (month/week/day). Internal consistencies for the adapted LSNS-6 subscale (n = 20) ranged from excellent to poor, respectively (i.e., 0.89/0.70/0.58 [month/week/day]).
Two staff members from each unit were shown a list of names and photographs of residents from their unit, and asked to identify friendships between these residents, and rate relationship strength as true (+3) or casual (+2) friendship. Positive relationships not identified as “friendship” were rated positive regard (+1). Staff were asked which of these residents were in conflict, indicating negative relationships (de Medeiros, Saunders, Doyle, Mosby, & Van Haitsma, 2011), and to rate relationship strength as mild disregard (−1), moderate dislike (−2), or strong dislike (−3). Staff were not asked but spontaneously reported positive, but no negative, between-unit relationships.
Network data were analyzed using UCINET 6 (Borgatti et al., 2002) and IBM SPSS 22.0. Within each unit, if staff reported discrepant perceptions on relationship presence and strength, relationship presence and weaker strength were selected. There were no discrepancies in relationship quality (positive or negative). Relationship dyads were separated into positive and negative, classified by strength (i.e., true friend, casual friend, positive regard; disregard, moderate dislike, strong dislike), and categorized by direction (i.e., was the relationship reciprocal; if not, who initiated the relationship/who received their overtures). Data were analyzed at the care unit level. Density was calculated as the proportion of dyadic relationships reported divided by the number of possible dyadic relationships. Residents' personal network size was calculated as the number of co-residents with whom the resident had direct relationships (Borgatti et al., 2002). Graphs generated in NodeXL 18.104.22.168 included reported between-unit relationships. Data were checked for normality in SPSS 22.0. Shapiro-Wilk tests indicated non-normal distributions for SNA data; nonparametric tests were used for analyses.
Characteristics of the 36 interviewed residents are summarized and reported in detail elsewhere (Casey et al., 2015). Participants ranged in age from 63 to 94 (mean age = 81.8 years), most were women (n = 22, 61.1%), and 41.7% (n = 15) were married. Time in care ranged from 0.25 to 10.75 years (median = 1.5 years). Twenty-seven residents had dementia. Barthel Index scores ranged from 0 to 75 (median score = 10, interquartile range [IQR] = 5 to 25). MOSES subscale (median score = 18, IQR = 14 to 24), adapted LSNS-6 subscale (median score = 5, IQR = 2 to 10), and Friendship Scale (median score = 14, IQR = 9 to 17) scores indicated residents were moderately engaged but self-reported moderate-to-high levels of perceived social isolation (Table).
Participant Characteristics and Median Scores on Barthel Index, Moses, LSNS-6, and Friendship Scales
Relationships between co-residents formed positive and negative networks (Figure). Of 91 residents, staff identified 52 (57.1%) initiating positive relationships (including between-unit relationships) and 24 (26.4%) initiating negative relationships. Approximately one third of residents initiated (n = 30, 33%) or received (n = 34, 37.4%) friendship (i.e., true and casual), and one third (n = 30, 33%) were isolates (i.e., having no relationships). The median size of positive relationship networks was 2 (IQR = 1 to 4), and the median size of negative networks was 2 (IQR = 1 to 3). Friendship network densities were low (<0.01 to 0.02), and densities for positive and negative networks were low overall (0.04 to 0.07 and 0.01 to 0.07, respectively). Residents' positive-to-negative network size ratios indicated median ratios of 1.5:1 for Unit 1 (n = 11, IQR = 1 to 5), 0:1 for DSU (n = 15, IQR = 0 to .5), and 0.7:1 for Unit 3 (n = 12, IQR = 0.2 to 2.3).
Staff-rated resident positive (top) and negative (bottom) social networks. Spheres represent all residents in each care unit (clockwise from lower left corner of each graph): Unit 1, n = 40; Unit 3, n = 33; Dementia-Specific Unit, n = 18. Arrows indicate tie direction. Line colors indicate relationship rating: dark green = true friend, light green = casual friend, white = positive regard; and yellow = mild disregard, orange = moderate dislike, red = strong dislike.
Number of positive and negative relationships and isolate status were correlated with social engagement (MOSES subscale; n = 36), perceived social support (Friendship Scale; n = 25), and objective social isolation (adapted LSNS-6 subscale; n = 22). Higher MOSES Withdrawn Behavior subscale scores correlated strongly with fewer reciprocated positive relationships (ρ = −0.71, p < 0.001), moderately with total negative relationships (ρ = 0.51, p = 0.002) and received not reciprocated negative relationships (ρ = 0.41, p = 0.012), and moderately with isolate status (ρ = 0.54, p = 0.001). Higher Friendship Scale scores correlated moderately with higher total positive (ρ = 0.44, p = 0.03) and reciprocated positive (ρ = 0.41, p = 0.044) relationships. Residents reporting higher objective isolation had more negative relationships (ρ = −0.44, p = 0.042), and residents reporting higher perceived social isolation received negativity in relationships with more co-residents without reciprocating (ρ = −0.41, p = 0.041). Higher perceived and objective social isolation correlated moderately and strongly with isolate status, respectively (ρ = −0.42, p = 0.038; ρ = −0.60, p = 0.003).
One third of residents had positive relationships with co-residents. One third of residents had no positive relationships. Previous research reports similar proportions of close relationships and loners among NH residents (Abbott, Bettger, Hampton, & Kohler, 2013; Retsinas & Garrity, 1985). Approximately one third of residents had negative relationships. Negative SNA data have not been reported previously for NH residents, but results are consistent with studies suggesting resident-to-resident aggression and negative relationships are common (Ferrah et al., 2015).
Staff-report of co-resident positive and negative relationships represented global perceptions of relationship quality rather than counts of affective experiences (Meeks et al., 2012). However, low positive-to-negative network ratios may have reflected residents' vulnerability to relationship-based negative affect. Residents' moderate levels of engagement were similar to those reported in previous research (Low et al., 2013). Residents reported as having more positive co-resident relationships were perceived as more socially engaged and felt more supported.
Study limitations include a cross-sectional design within one facility; therefore, results may reflect effects specific to the moment or facility. Correlational analyses were powered only to detect strong statistically significant relationships. Few staff members were interviewed, providing limited insight into overall staff perceptions. Residents who agreed to interviews may not represent the total facility population; non-participants may have been even more isolated.
Findings highlight NH residents' isolation and lack of engagement. Given the few friendships, common negative relationships, and low positive-to-negative network ratios, monitoring and cultivating co-resident relationships may be important to creating a social environment that supports the person-hood and well-being of residents with dementia. PCC should include greater attention to the broad social context in which residents live and the quality of co-resident relationships. Staff cannot control affection between residents but can address negative interactions and initiate and promote positive interactions among those who cannot do so themselves.
The current study is the first to demonstrate NH residents' negative social networks using SNA data. Residents' level of social engagement and experiences of support or isolation were associated with the quantity and quality of their relationships with co-residents. Further research may determine whether relationship improvement translates to improvement in residents' social engagement and perceived support.
- Abbott, K.M., Bettger, J.P., Hampton, K.N. & Kohler, H.-P. (2013). The feasibility of measuring social networks among older adults in assisted living and dementia special care units. Dementia, 14, 199–219. doi:10.1177/1471301213494524 [CrossRef]
- Ashida, S. & Heaney, C.A. (2008). Differential associations of social support and social connectedness with structural features of social networks and the health status of older adults. Journal of Aging and Health, 20, 872–893. doi:10.1177/0898264308324626 [CrossRef]
- Borgatti, S., Everett, M. & Freeman, L. (2002). Ucinet for Windows: Software for social network analysis (Version 6.523). Harvard, MA: Analytic Technologies. Retrieved from https://sites.google.com/site/ucinetsoftware/downloads
- Cacioppo, S., Capitanio, J.P. & Cacioppo, J.T. (2014). Toward a neurology of loneliness. Psychological Bulletin, 140, 1464–1504. doi:10.1037/a0037618 [CrossRef]
- Casey, A.-N.S., Low, L.-F., Jeon, Y.-H. & Brodaty, H. (2015). Residents' perceptions of friendship and positive social networks within a nursing home. The Gerontologist. Advance online publication. doi:10.1093/geront/gnv146 [CrossRef]
- de Medeiros, K., Saunders, P.A., Doyle, P.J., Mosby, A. & Van Haitsma, K. (2011). Friendships among people with dementia in long-term care. Dementia, 11, 363–381. doi:10.1177/1471301211421186 [CrossRef]
- English, T. & Carstensen, L.L. (2014). Selective narrowing of social networks across adulthood is associated with improved emotional experience in daily life. International Journal of Behavioral Development, 38, 195–202. doi:10.1177/0165025413515404 [CrossRef]
- Ferrah, N., Murphy, B.J., Ibrahim, J.E., Bugeja, L.C., Winbolt, M., LoGiudice, D. & Ranson, D.L. (2015). Resident-to-resident physical aggression leading to injury in nursing homes: A systematic review. Age and Ageing, 44, 356–364. doi:10.1093/ageing/afv004 [CrossRef]
- Golden, J., Conroy, R.M. & Lawlor, B.A. (2009). Social support network structure in older people: Underlying dimensions and association with psychological and physical health. Psychology, Health & Medicine, 14, 280–290. doi:10.1080/13548500902730135 [CrossRef]
- Hawthorne, G. (2006). Measuring social isolation in older adults: Development and initial validation of the Friendship Scale. Social Indicators Research, 77, 521–548. doi:10.1007/s11205-005-7746-y [CrossRef]
- Helmes, E., Csapo, K.G. & Short, J.-A. (1987). Standardization and validation of the Multidimensional Observation Scale for Elderly Subjects (MOSES). Journal of Gerontology, 42, 395–405. doi:10.1093/geronj/42.4.395 [CrossRef]
- Kitwood, T. (1993). Towards a theory of dementia care: The interpersonal process. Ageing & Society, 13, 51–67. doi:10.1017/S0144686X00000647 [CrossRef]
- Low, L.-F., Brodaty, H., Goodenough, B., Spitzer, P., Bell, J.-P., Fleming, R. & Chenoweth, L. (2013). The Sydney Multisite Intervention of LaughterBosses and ElderClowns (SMILE) study: Cluster randomised trial of humour therapy in nursing homes. BMJ Open, 3, e002072. doi:10.1136/bmjopen-2012-002072 [CrossRef]
- Lubben, J., Blozik, E., Gillmann, G., Iliffe, S., von Renteln Kruse, W., Beck, J.C. & Stuck, A.E. (2006). Performance of an abbreviated version of the Lubben Social Network Scale among three European community-dwelling older adult populations. The Gerontologist, 46, 503–513. doi:10.1093/geront/46.4.503 [CrossRef]
- Mahoney, F.I. & Barthel, D. (1965). Functional evaluation: The Barthel Index. Maryland State Medical Journal, 14, 56–61.
- Meeks, S., Van Haitsma, K., Kostiwa, I. & Murrell, S.A. (2012). Positivity and well-being among community-residing elders and nursing home residents: What is the optimal affect balance?Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 67, 460–467. doi:10.1093/geronb/gbr135 [CrossRef]
- Retsinas, J. & Garrity, P. (1985). Nursing home friendships. The Gerontologist, 25, 376–381. doi:10.1093/geront/25.4.376 [CrossRef]
- Stein-Parbury, J., Chenoweth, L., Jeon, Y.H., Brodaty, H., Haas, M. & Norman, R. (2012). Implementing person-centered care in residential dementia care. Clinical Gerontologist, 35, 404–424. doi:10.1080/07317115.2012.702654 [CrossRef]
Participant Characteristics and Median Scores on Barthel Index, Moses, LSNS-6, and Friendship Scales
|Variable||Unit 1 (n = 15)||DSU (n = 9)||Unit 3 (n = 12)||Total (N = 36)|
|Age (years) (mean, SD)||82.9 (7.7)||83.2 (5.5)||79.3 (9)||81.8 (7.7)|
|Years in care (median, 1st to 3rd quartile)||1 (0.5 to 1.9)||2 (0.7 to 5.8)||2 (0.4 to 2.9)||1.5 (0.5 to 2.7)|
|Women (n, %)||7 (46.7)||7 (77.8)||8 (66.7)||22 (61.1)|
|Married/partnered (n, %)||6 (40)||2 (22.2)||7 (58.3)||15 (41.7)|
|Dementia (n, %)||10 (66.7)||9 (100)||8 (66.7)||27 (75)|
|Barthel Index scorea (median, 1st to 3rd quartile)||15 (5 to 35)||5 (0 to 17)||10 (1 to 18)||10 (5 to 25)|
|MOSES withdrawn behavior subscale scoreb (median, 1st to 3rd quartile)||16 (12 to 17)||27 (19 to 28)||21 (16 to 25)||18 (14 to 24)|
|Adapted LSNS-6 friendship subscale scorec (median, 1st to 3rd quartile)||5 (3 to 11)||0 (0 to 8)||8 (2 to 9)||5 (2 to 10)|
|Friendship Scale scored (median, 1st to 3rd quartile)||17 (10 to 20)||7 (6 to 9)||14 (11 to 16)||14 (9 to 17)|